PLoS ONE (Jan 2011)
Age determination by back length for African savanna elephants: extending age assessment techniques for aerial-based surveys.
Abstract
Determining the age of individuals in a population can lead to a better understanding of population dynamics through age structure analysis and estimation of age-specific fecundity and survival rates. Shoulder height has been used to accurately assign age to free-ranging African savanna elephants. However, back length may provide an analog measurable in aerial-based surveys. We assessed the relationship between back length and age for known-age elephants in Amboseli National Park, Kenya, and Addo Elephant National Park, South Africa. We also compared age- and sex-specific back lengths between these populations and compared adult female back lengths across 11 widely dispersed populations in five African countries. Sex-specific Von Bertalanffy growth curves provided a good fit to the back length data of known-age individuals. Based on back length, accurate ages could be assigned relatively precisely for females up to 23 years of age and males up to 17. The female back length curve allowed more precise age assignment to older females than the curve for shoulder height does, probably because of divergence between the respective growth curves. However, this did not appear to be the case for males, but the sample of known-age males was limited to ≤27 years. Age- and sex-specific back lengths were similar in Amboseli National Park and Addo Elephant National Park. Furthermore, while adult female back lengths in the three Zambian populations were generally shorter than in other populations, back lengths in the remaining eight populations did not differ significantly, in support of claims that growth patterns of African savanna elephants are similar over wide geographic regions. Thus, the growth curves presented here should allow researchers to use aerial-based surveys to assign ages to elephants with greater precision than previously possible and, therefore, to estimate population variables.